Absolute velocity damping algorithm with varying damping ratio for inertial navigation systems based on Kalman filter

The navigation information of an inertial navigation system (INS) is contaminated by period oscillating errors of which the amplitudes are invariant or increasing with time. A velocity damping algorithm based on Kalman filter is proposed to attenuate the oscillating errors in this paper. The differences between external velocity measurements, e.g. GPS or odometer velocity measurements, and INS velocity outputs are applied as control signals to add damping to the system. To find the optimal feedback gains for the control signals, an imaginary system model is developed according to the criterion that the estimation error vector of the states by Kalman filter is identical to the error vector of INS. Thus the elements in the gain matrix of Kalman filter are exactly the optimal gains to be determined. The horizontal velocity damping mechanization is first introduced; followed by the more complex one, i.e. absolute velocity damping mechanization. Simulations are presented in two sections respectively to illustrate the effectiveness of the above two mechanizations. The proposed method breeds a time-variant damping ratio, endowing the system with sound transient and steady-state performance. Finally, the performance of absolute velocity damping algorithm is verified on a dual-axis-rotation strapdown INS with GPS serving as the source of reference velocity.